New Research Shows Promise and Limitations of Physicians Working with GPT-4 for Decision Making

Published in JAMA Network Open, a collaborative team of researchers from the University of Minnesota Medical School, Stanford University, Beth Israel Deaconess Medical Center and the University of Virginia studied how well doctors used GPT-4 - an artificial intelligence (AI) large language model system - for diagnosing patients.

The study was conducted with 50 U.S.-licensed physicians in family medicine, internal medicine and emergency medicine. The research team found that the availability of GPT-4 to physicians as a diagnostic aid did not significantly improve clinical reasoning compared to conventional resources. Other key findings include:

  • GPT-4 alone demonstrated significantly better scores in diagnostic performance, surpassing the performance of clinicians using conventional diagnostic online resources and clinicians assisted by GPT-4.
  • There was no significant enhancement in diagnostic performance with the addition of GPT-4 when assessing clinicians using GPT-4 against clinicians using conventional diagnostic resources.

"The field of AI is expanding rapidly and impacting our lives inside and outside of medicine. It is important that we study these tools and understand how we best use them to improve the care we provide as well as the experience of providing it," said Andrew Olson, MD, a professor at the U of M Medical School and hospitalist with M Health Fairview. "This study suggests that there are opportunities for further improvement in physician-AI collaboration in clinical practice."

These results underline the complexity of integrating AI into clinical practice. While GPT-4 alone showed promising results, the integration of GPT-4 as a diagnostic aid alongside clinicians did not significantly outperform the use of conventional diagnostic resources. This suggests a nuanced potential for AI in healthcare, emphasizing the importance of further exploration into how AI can best support clinical practice. Further, more studies are needed to understand how clinicians should be trained to use these tools.

The four collaborating institutions have launched a bi-coastal AI evaluation network - known as ARiSE - to further evaluate GenAI outputs in healthcare.

Funding for this research was provided by the Gordon and Betty Moore Foundation.

Goh E, Gallo R, Hom J, Strong E, Weng Y, Kerman H, Cool JA, Kanjee Z, Parsons AS, Ahuja N, Horvitz E, Yang D, Milstein A, Olson APJ, Rodman A, Chen JH.
Large Language Model Influence on Diagnostic Reasoning: A Randomized Clinical Trial.
JAMA Netw Open. 2024 Oct 1;7(10):e2440969. doi: 10.1001/jamanetworkopen.2024.40969

Most Popular Now

Open Medical Works with Moray's Dig…

Open Medical is working with the Digital Health & Care Innovation Centre’s Rural Centre of Excellence on a referral management plan, as part of a research and development scheme to...

Generative AI on Track to Shape the Futu…

Using advanced artificial intelligence (AI), researchers have developed a novel method to make drug development faster and more efficient. In a new paper, Xia Ning, lead author of the study and...

Reorganisation, Consolidation, and Cuts:…

NHS England has been downsized and abolished. Integrated care boards have been told to change function, consolidate, and deliver savings. Trusts are planning big cuts. The Highland Marketing advisory board...

AI Tool Uses Face Photos to Estimate Bio…

Eyes may be the window to the soul, but a person's biological age could be reflected in their facial characteristics. Investigators from Mass General Brigham developed a deep learning algorithm...

Philips Future Health Index 2025 Report …

Royal Philips (NYSE: PHG, AEX: PHIA), a global leader in health technology, today unveiled its 2025 Future Health Index U.S. report, "Building trust in healthcare AI," spotlighting the state of...

Personalized Breast Cancer Prevention No…

A new telemedicine service for personalised breast cancer prevention has launched at preventcancer.co.uk. It allows women aged 30 to 75 across the UK to understand their risk of developing breast...

New App may Help Caregivers of People Ge…

A new study by investigators from Mass General Brigham showed that a new app they created can help improve the quality of life for caregivers of patients undergoing bone marrow...

An App to Detect Heart Attacks and Strok…

A potentially lifesaving new smartphone app can help people determine if they are suffering heart attacks or strokes and should seek medical attention, a clinical study suggests. The ECHAS app (Emergency...

A Machine Learning Tool for Diagnosing, …

Scientists aiming to advance cancer diagnostics have developed a machine learning tool that is able to identify metabolism-related molecular profile differences between patients with colorectal cancer and healthy people. The analysis...

Fine-Tuned LLMs Boost Error Detection in…

A type of artificial intelligence (AI) called fine-tuned large language models (LLMs) greatly enhances error detection in radiology reports, according to a new study published in Radiology, a journal of...

DeepSeek-R1 Offers Promising Potential t…

A joint research team from The Hong Kong University of Science and Technology and The Hong Kong University of Science and Technology (Guangzhou) has published a perspective article in MedComm...

Deep Learning can Predict Lung Cancer Ri…

A deep learning model was able to predict future lung cancer risk from a single low-dose chest CT scan, according to new research published at the ATS 2025 International Conference...